People currently align connectors to modules but typically do it via either an active optical alignment scheme (where they emit light into or from individual devices) or use very small numbers of devices where an accurate pick & place machine can get integration alignment. For example, using one laser where there is no concern regarding rotational alignment.
The processes typically used for alignment of connectors requires the individual devices be illuminated and then the fiber(s) are scanned across the optical device with the output light from the end of the fiber monitored for the intensity of light output. This process is repeated and the fiber light output is continuously monitored as fibers are moved in several dimensions to allow accurate alignment. An example of this technique is described in, for example, IBM Micro News, Volume 6, Number 3, Third Quarter 2000.
Such techniques are costly, since requiring illuminating devices necessitates the use of significant capital equipment to power up each device, to monitor the output powers, etc. Moreover, because the techniques are active device techniques, they run the risk of damaging the devices.